There is recent evidence that marijuana use has increased, and that a subset of marijuana smokers are seeking treatment for their marijuana abuse. These data, in combination with empirical evidence that marijuana dependence develops in humans, support the objective of this proposal to develop a model to assess the ability of pharmacological interventions to decrease marijuana use. Specifically, the acute and residual effects of smoked marijuana under both an active and placebo dose of study medication will be assessed. The underlying assumption of this proposal is that marijuana self-administration can be disrupted via several mechanisms. The following behaviors will be measured: marijuana self- administration, psychomotor task performance, social behavior, food intake, and subjective reports of drug effects. Aim #1. Evaluate the ability of the anticonvulsant, divalproex, and the antidepressant, bupropion to decrease marijuana use indirectly by attenuating symptoms of smoked marijuana withdrawal. Hypotheses: these medications will I) attenuate symptoms of irritability, anxiety, and depression during marijuana abstinence, and 2) correspondingly, reduce marijuana self-administration after a period of monitored marijuana abstinence. Aim #2. Evaluate the ability of the mu-opioid antagonist, naltrexone to directly decrease marijuana self-administration by decreasing its acute effects. Hypotheses: naltrexone will 1) decrease the subjective and behavioral effects of smoked marijuana, 2) decrease marijuana self- administration, and 3) decrease the development of marijuana dependence, evidenced by fewer symptoms of abstinence. Thus, naltrexone may be a useful treatment medication in two ways: by decreasing both marijuana smoking and the development of marijuana dependence. Although smoked marijuana is one of the most widely abused drugs in the world, little is known about effective treatments for marijuana abuse and dependence. The strength of this protocol lies in the utilization of a controlled laboratory setting to examine the interactive effects of potential treatment medications with marijuana self-administration and a range of its behavioral effects. The data collected will suggest more efficacious approaches to treating marijuana abusers.